OPTICAL VIBRATION MEASUREMENT ON DISTANTLY SITUATED OBJECTS

Citation
W. Georgi et al., OPTICAL VIBRATION MEASUREMENT ON DISTANTLY SITUATED OBJECTS, TM. Technisches Messen, 60(6), 1993, pp. 223-227
Citations number
NO
Categorie Soggetti
Instument & Instrumentation
Journal title
ISSN journal
01718096
Volume
60
Issue
6
Year of publication
1993
Pages
223 - 227
Database
ISI
SICI code
0171-8096(1993)60:6<223:OVMODS>2.0.ZU;2-D
Abstract
To measure the vibration of distantly situated objects, such as a brid ge loaded by traffic, optical mehtods are used among other things. The y partially rest on observing a punctiform light source which has to b e fixed on a swinging object. In other cases it is already sufficient to notice a linear light-dark contrast, like given with a swinging ope n-air transmission line appearing in a telescope as a black line in fr ont of the bright background of the sky. But all these methods are ver y sensitive to optical inhomogeneities which may develop on hot days d ue to rising airflows. In contrast to that the following optical metho d is largely insensitive to optical inhomogeneities. Its basic idea is to photograph the swinging object several times one after the other. In doing so the scan frequency has to be higher than the frequency of the observed vibration. From the shiftings of the picture on the linea r CCD sensor of the camera the vibration can be reconstructed. It is p ossible to determine therefrom its amplitude and its frequency. The pr ocess of auto-correlation in three different versions as well as the ' 'Fast Fournier Transformation'' (FFT) were used to evaluate the pictur es. The results of all 4 methods were satisfactory, but it turned out that the process of correlation leads to the most reliable and precise results. The experiments were evaluated by a commercial personal comp uter (80386 DX with 25 MHz clock-frequency). By increasing the perform ance of the computer 6 times, the fastest way, which is limited to har monic oscillations, would make it possible to process the datas just a s quick as the photographs were taken. The remaining methods can also be used for other kinds of vibration. They are 2 to 12 times slower. I n the following the 3 methods of correlation are discussed Explanation s concerning the FFT are reserved to a following article (part 2).